Environmental Stability

A TBC must also resist environmental degradation. This may occur as oxidation of the bond coat, phase transformation stress deterioration of the zirconia layer, or chemical destabilization of the oxide layer. The severity of these effects depends on the oxidation resistance of the bond coat, the initial crystalline phase mixture of the zirconia coating, the service temperature, and the environmental impurities present.

Bond Coat Oxidation. To test the oxidation resistance of the bond coat and its effect on the spallation of the TBC, a thermal cycle needs to combine extended soak time at high temperatures with periodic heating and cooling cycling to induce thermal stress. One such cycle used to test TBCs for aircraft gas turbine applications is a simple tube furnace that operates at a constant temperature of 1121 °C (2050 °F), with the sample set pneumatically moved in and out of the hot zone. The TBC samples rest on an alumina firebrick boat, with dense alumina pushrods made from thermocouple insulator ceramics, for an all-oxide hot-zone support system. The samples heat to furnace temperature in about 5 min, soak for 50 min, and then are quickly pulled out for a 5 min cooling period. At the end of every 25 cycles, the samples are held out and visually examined at room temperature. After 200 cycles, the samples are inspected for spallation or edge cracking, then are mounted in plastic for evaluation in polished cross section. There the extent of bond coat oxidation and interface cracking can be measured. The test temperature is as much as 110 °C (200 °F) higher than expected for the component in service and is thus an accelerated test for oxidation of the bond coat. To make this comparison test valid, care should be taken that the furnace temperature profile is uniform and covers the full zone occupied by the TBC samples.

A cycle for testing coatings for diesel applications can be similarly designed by setting the soak temperature about 110 °C (200 °F) higher than the measured steady-state temperature of the substrate metal of the component in question. In the case of coatings for steel or cast iron substrates, the bond coat is typically not heat treated, which may allow internal oxidation of the Ni-Cr or Fe-Cr-Al-Y alloys used for thermally sprayed layers. Thus, a cycle soak temperature of 650 to 982 °C (1200 to 1800 °F) may still challenge inadequate coating systems.

Zirconium Oxide Phase Stability. Pure zirconium oxide is an allotropic material and has a cubic structure from its freezing point at 2680 °C (4856 °F) down to 2370 °C (4298 °F), where it transforms to a tetragonal structure of similar cell size. Because of this small difference in crystalline cell volume, thermal cycling across the cubic-tetragonal transition temperature does not impose severe internal stresses. Following further cooling to below 1170 °C (2138 °F), zirconia transforms to the monoclinic structure, which has about a 4% larger crystalline cell volume than the tetragonal structure (Ref 3). Large internal stresses are generated as the transformation front sweeps through the material, which can lead to crack initiation. Commercially useful zirconia is alloyed with yttria, which has a cubic-plus-tetragonal two-phase field in its phase diagram (Fig. 6) and inhibits the low-temperature tetragonal-to-monoclinic transformation (Ref 3, 5, 6).

Fig. 6 Phase diagram of the ZrO2-rich region of the ZrO2-Y2O3 system. M, monoclinic phase; C, cubic phase; T, tetragonal phase; L, liquid; t, nontransformable tetragonal. Source: Ref 4

In thermally sprayed yttria-stabilized coatings, the feedstock is powder, which may be fabricated by a variety of methods (Ref 7). Partially stabilized zirconia, with about 6.5 to 9 wt% Y, mostly avoids the monoclinic phase and the fully stabilized cubic-only phase, which is known to have less thermal shock resistance than the dual-phase cubic-plus-tetragonal structure (Ref 8). A powder feedstock may have the correct average yttria content, but it may not be homogeneous at the crystalline grain level. An inhomogeneous powder may have some local volumes that have no yttria and others with yttria concentrations far higher than the bulk analysis. The x-ray diffraction pattern of such a powder thus shows the monoclinic phase characteristic of the nearly pure zirconia grains, along with tetragonal and cubic phases from the volume of the material that was sufficiently combined with yttria. When an inhomogeneous powder is plasma sprayed to form a coating, the particle melting in the arc improves the degree of homogeneity. However, this may not be sufficient to fully homogenize the material, depending on the state of the starting powder and the chosen spray parameters.

Figure 7 shows the x-ray diffraction patterns of an inhomogeneous powder and a sprayed coating. In this example, the spray conditions did eliminate nearly all of the starting monoclinic phase. The following table gives the phase distribution for the two materials in Fig. 7, based on the areas under the diffraction peaks:

Material

Phase, mol%

Monoclinic

Tetragonal

Cubic

Starting powder

6

76

18

Fig. 7 X-ray diffraction patterns of yttria-stabilized zirconia powder showing some monoclinic phase, and of a coating made from that powder. M, monoclinic phase; C, cubic phase; T, tetragonal phase; T + C, overlapping tetragonal and cubic reflection

The values in the table were obtained with the Miller algorithm (Ref 6):

where M, T, and C denote the mole percentages of the monoclinic, tetragonal, and cubic phases, respectively, and 7(hkl) is the integrated intensity for the (hkl) diffraction peak. For the high-angle (400) peaks, a deconvolution routine was used first to separate the three overlapping peaks.

Not all TBCs need to be made from homogeneous 6 to 9 wt% YSZ. The importance of the degree of homogeneity and the allowable amount of monoclinic phase in the starting powder depend on the maximum temperature to be experienced by the TBC. Consider the zirconia-yttria phase diagram in Fig. 6 in the following argument. Assume that a coating made from inhomogeneous 8% YSZ powder has some fraction of grains of very low yttria content. These grains will behave as if they were a separate entity and experience the phase transitions characteristic of nearly pure zirconia. At room temperature these grains should exist in the monoclinic structure. The other grains may be partially stabilized with yttria and exist mainly as the nontransformable tetragonal phase, and possibly some cubic-phase material. These higher-yttria-content nontransformable tetragonal grains exist as a nonequilibrium phase because of the rapid quench from the liquid state upon deposition at the substrate. They have high thermal fatigue resistance because they will not transform to the monoclinic phase and experience the large volume change. On the other hand, the nearly pure zirconia grains having the monoclinic structure may be safe from transformation stresses as long as they remain below the temperature of the M + T two-phase field, which can be as high as about 1100 °C (2000 °F).

Therefore, coated components that have a maximum exposure of only 760 °C (1400 °F) should be stable against deleterious phase transformation, even though they may have a zirconia coating composed of both stabilized and unstabilized grains. As the operating temperature of the component increases, the need for a homogeneous coating increases. Applications at temperatures above about 1100 °C (2000 °F) should require fully homogeneous coatings. For applications at 760 to 1100 °C, thermal cycle testing is needed to determine whether the powder and coating process selected are adequate. If monoclinic-phase material exists in the coating and is tested with flame impingement on the oxide face to temperatures that cycle through the transformation temperature, the surface layer will begin to flake away small particles. Because of the insulating effect of zirconia, the lower layers of the coating would be at a lower temperature and thus remain intact. If the same nonhomogeneously stabilized coating is exposed to a uniform temperature test, the whole zirconia layer could develop cracks, leading to early spallation if the whole layer is cycled through the phase-transformation temperature.

Although the use of fully homogeneously stabilized coatings may seem warranted for all applications, regardless of operating temperature, economic considerations argue against this. Fully homogeneous powders, which are more costly, must either be fused and crushed or sintered from pure zirconia and yttria components at high temperatures. If the exposure temperature of the TBC does not demand monoclinic-phase-free material, a less costly powder will still meet the thermal barrier performance required. The key is to know the relation between operating temperature limits and the phase content of the coating.

Chemical Effects on Phase Stability. As discussed above, powders that are inhomogeneous in regard to yttria concentration could lead to coatings with phase-transformation-induced cracking if used at high temperatures. It has also been found that fully homogeneous yttria-stabilized coatings used in high-temperature environments containing vanadium and/or sulfur can experience surface degradation. In this case, the impurities can form yttrium vanadate and yttrium sulfide by leaching yttrium from the coating (Ref 5, 9). This leads to a surface layer that is progressively depleted in yttria content, and eventually to a material that is in the fully monoclinic structure when cooled and has a large volume change and disruptive stresses when thermally cycled. This surface transformation can also be detected by x-ray diffraction, which will give the evidence needed to understand the failure mechanism.

References cited in this section

3. J.R. VanValzah and H.E. Eaton, Cooling Rate Effects on the Tetragonal to Monoclinic Phase Transformation in Aged Plasma-Sprayed Yttria Partially Stabilized Zirconia, Surf. Coat. Technol., Vol 46, 1991, p 289-300

4. H.G. Scott, Phase Relationships in the Zirconia-Yttria System, J. Mater. Sci., Vol 10 (No. 9), 1975, p 15271535

5. R.J. Bratton and S.K. Lau, Zirconia Thermal Barrier Coatings, Adv. Ceram., Vol. 3: Science and Technology of Zirconia, American Ceramic Society, 1981, p 226-240

6. E.C. Subbarao, Zirconia --An Overview, Adv. Ceram., Vol. 3: Science and Technology of Zirconia, American Ceramic Society, 1981, p 1-13

7. M.F. Gruninger and M.V. Boris, Thermal Barrier Ceramics for Gas Turbine and Reciprocating Heat Engine Applications, Proc. ITSC, C.C. Berndt, Ed., ASM International, 1992, p 487-492

8. R.A. Miller, J.L. Smialek, and R.G. Garlick, Phase Stability in Plasma-Sprayed Partially Stabilized Zirconia-Yttria, Adv. in Ceram., Vol. 3: Science and Technology of Zirconia, American Ceramic Society, 1981, p 241253

9. B.A. Nagaraj and D.J. Wortman, Burner Rig Evaluation of Ceramic Coatings with Vanadium Contaminated Fuel, J. Eng. Gas Turbines Power, Vol 112, 1990, p 536-542

0 0

Post a comment